The N-G ratios are for fleets of cars they don’t get into the individual makeup of the cars within the fleet numbers.  

Randy, all,

"Understanding your road's traffic is very important" can't be emphasized enough. Bruce mentioned Reading traffic in his PRR area of interest (Columbia, PA, about 20 miles south of Harrisburg on the east side of the Susquehanna river.)

My area of interest is the mainline starting about 30 miles RR west of Columbia. Purchase of several WWII vintage PRR documents on car routing provides some major insights. Yes, for loads, many of us have decided that the distribution of loaded cars can match Bruce's proposed numbers, especially for a mainline functioning as a major US bridge line (as the PRR was between Pittsburgh and east coast destinations, especially during WWII.)

One of the PRR documents I found was specifically on how to route foreign road empties towards their home roads. Keep in mind on the PRR during WWII, there was a tremendous imbalance of loads versus empties. Less than 1% empties EB, 75% empties WB. Ignoring the PRR's large EB coal traffic, 60% of the WB non-hopper cars were empty.

While the 40% of those cars that were loaded might still reflect the national averages, for the empties, there could be significant deviations. Case in point was the NYC - operator of the second largest fleet of box cars. If the yardmaster's followed the PRR guidance, there would be no WB empty NYC cars west of Harrisburg on the PRR mainline. They were sent in a different direction (in fact in central PA, the PRR seems to have enjoyed returning NYC empties to a location on the NYC that was likely very inconvenient for the NYC to deal with ;-)  The NYC cars would still be used for loads, but for WB non-hopper trains, the NYC cars should be under-represented.

The other twist was limitations on car clearances, which in the east were often inadequate for many western RR XM's.  For the PRR, that meant that the taller XM's would be on trains running between Harrisburg and Chicago, since there were clearance limits along the line towards St. Louis. (The PRR booklet lists every single class of XM for every railroad, with notes on where they could, and could not, be routed.)

While not impacting the average distribution of cars flowing across the railroad, it could have significant impact on the consists of individual merchandise trains, just because of their ultimate destination, For example, all the tall XM's (loads and MTY's) on WB trains were routed towards Chicago (for tall cars destined for the SW US, those cars would use a division in Indiana to get back to the line to St. Louis). XM's on the trains headed directly toward St. Louis would be limited to the shorter XM's.  This offers some unique challenges a superintendent may want to impose on their yardmaster when classifying and blocking XM's....

Dave Evans

--
Tim O'Connor
Sterling, Massachusetts

The G-N or N-G data really only works up to the mid 1950s. This is well stated in the analyses done by Tim and Dave.  Once you get later things changed, piggyback, specialized boxcars, specific service, older cars on the books but stored all over the place (looking at you Pennsy), etc. 

I've transcribed 10,000 car movements from 1960's train list for the Chicago Great Western and the home road cars only made up 10% of the total.

Along with that, in the top 10 were Canadian Pacific and Canadian National.

Jason Kliewer

Although I generally agree with the Gilbert-Nelson approach, along with Bruce Smith’s percentages, they are very dependent upon the road you model. 

Bruce models the PRR, so 75% home road hoppers on any coal hauling road makes sense. But I model the NH that has a modest fleet of hoppers of less than 1,000, with little originating traffic that would send them offline. So home road hoppers make up a very small part of my fleet.

Understanding your road’s traffic is very important. For example, as much as 60% of the coal entering CT in my era was by water. That appears to have been Pocahontas and other bituminous coming through Baltimore, and I think was the primary source of coal loaded in NH hoppers to be distributed to industry, but particularly the railroad’s facilities. On the other hand, anthracite came primarily via Maybrook, unless you were in the vicinity of NYC, in which case it came mostly by float via Jersey City and Greenville. So I go with the anthracite hoppers as a mix of 75% home road of their loading point.

So the mix of hoppers is very dependent on whether you’re modeling a coal road or not, I think.

Gondolas and flats are the same, and since the NH was heavily involved in manufacturing, gondolas were also predominantly foreign. Like the hoppers, I think gondolas and flats are far more dependent on the home road of the loading point. 

I also think we often have too little variety (and too few) “rare” cars. I have two pictures, from different days, with ATSF Ga-8 gons, and these weren’t photographed because of their rarity, they just happened to be on those trains. Likewise the photos of the Litchfield & Madison gondola in New Britain. Numerous pictures of C&I hoppers, etc.

I’m not saying those happen every day, but based on those and many, many other photos that have “rare” cars (often in the background) show that there’s a wider variety than many allow. I think we often make the mistake of not having enough different “rare” cars to mix in, making the couple of rare cars that are owned very common on our layouts.

Randy
--
—
Randy Hammill
Modeling the New Haven Railroad 1946-1954
http://newbritainstation.com

Like most people, I have far more rolling stock than will fit onto my model railroad.  To determine how many freight cars to place on the tracks I calculated the capacity of all sidings - yard, passing and industrial based on 40' cars - and divided that in half to set the maximum number of cars to be on the railroad at one time.  My model railroad is based on an NYC single-track branch line to a coal mine so I have more NYC cars than other roads.  Also, I have no refrigerator cars because I have no on-line industry that would use reefers.  The coal mine has a capacity of nine two-bay hoppers so I need only eighteen hoppers.  I have one retail coal yard so I need two hoppers from roads that serviced anthracite mines - one load and one empty.  I used information that I found years ago in magazine articles to determine how many foreign road cars to have on the roster.  It all worked out well because my model railroad is not clogged with cars, there is no gridlock anywhere.  Like prototype railroads, my model railroad has lots of empty track which I think enhances the realism.  Too many model railroads have too much track and way too many cars and locomotives.  On my model railroad there are four NYC GP-7 locomotives and four NYC cabooses.  Passenger service was discontinued.  The era is 1953 with no rolling stock or vehicles any newer than 1953.  Discussing ratios is an excellent topic, by the way!!!

Armand,

I would like to open the topic of ratios and how they might be used to develop a realistic roster for the specific era,road and locations Armand Premo